Systematic Study of Proton, Two-Proton, Alpha, and Cluster Radioactivity Half-Lives based on the Deformed Gamow-like Model and Tabular Prior-data Fitted Network (TabPFN)

arXiv:2609.21646 · nucl-th · Submitted 2026-09-18 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Nuclear Theory. The day in nuclear theory: structure, reactions and the models behind them.

Kai: I'm Kai.

Mira: And I'm Mira, here with Lev.

Kai: We're looking at the paper: "Systematic Study of Proton, Two-Proton, Alpha, and Cluster Radioactivity Half-Lives based on the Deformed Gamow-like Model and Tabular Prior-data Fitted Network (TabPFN)".

Mira: Here's the gist of the paper. A hybrid framework combining the deformed Gamow-like model (DGLM) with the Tabular Prior-data Fitted Network (TabPFN) is developed to improve half-life predictions for two-proton emission, proton emission, α decay, and cluster radioactivity Model Performance Comparison The DGLMb + TabPFN model achieves the best overall performance with a sigma RMS of 0.423, corresponding to an improvement of approximately 82.2% over DGLMb.

Kai: After introducing TabPFN, the prediction errors for proton emission and α decay are reduced by approximately 80.6% and 87.6%, respectively.

Mira: The model also reproduces the systematic evolution of α-decay half-lives and the shell-closure effect around N = 126.

Kai: Theoretical Frameworks The DGLM framework relates the half-lives T1/2 to the decay constant λ through T1/2 = ln 2 / λ × 10h.

Title and authors: Kai: So we're looking at this paper today titled "Systematic Study of Proton, Two-Proton, Alpha, and Cluster Radioactivity Half-Lives based on the Deformed Gamow-like Model and Tabular Prior-data Fitted Network (TabPFN)".

Mira: That title is pretty descriptive. It tells us we're looking at half-lives for several different types of radioactive decay—proton emission, two-proton emission, alpha decay, and cluster radioactivity.

Lev: It sounds like they are trying to unify a few different ways of thinking about these decays using this new framework.

Kai: Exactly. The main thing here is they're combining the Deformed Gamow-like Model with something called the Tabular Prior-data Fitted Network, or TabPFN.

Mira: That's the core idea. They’re taking an old physical model and hooking it up to a machine learning tool to make predictions much better for these specific decay types.

Lev: So it’s not just one thing; it’s a hybrid approach designed for those specific decay modes.

The paper's summary: Kai: Okay, so what the paper actually does in terms of the summary is they investigate a total of five hundred eighty-three radioactive nuclei across four decay types <ref:2609.21646#pg1,a total of 583 radioactive nuclei>.

Mira: They have seventeen two-proton emitters, forty-two proton emitters, and then they have a pretty large sample of four hundred ninety-eight alpha emitters and twenty-six cluster emitters <ref:2609.21646#pg1,17 two-proton emitters, 42 proton emitters>.

Lev: So it’s a systematic study covering a good range of isotopes to test the model across all these decay channels.

Kai: And the main takeaway from their summary is that this DGLMb plus TabPFN approach achieves the best overall performance among four models they compared.

Mira: They found that this combined model, DGLMb plus TabPFN, has a sigma RMS of zero point four two three, which they say is an improvement of approximately eighty-two point two percent over just using the DGLMb model alone <ref:2609.21646#pg1,an improvement of approximately 82.2% over>.

Lev: That eighty-two percent improvement sounds substantial when you're talking about predictive accuracy in nuclear physics <ref:2609.21646#pg1>.

The paper's improvements: Kai: The paper points out a few specific improvements they made to the original framework.

Mira: They focus on how this new TabPFN component helps refine the predictions for different decay modes.

Lev: I see they specifically mention that for proton emission and alpha decay, the prediction errors are reduced by about eighty point six percent and eighty-seven point six percent respectively <ref:2609.21646#pg1>.

Kai: That's a big jump in accuracy for those two specific channels, which is what the paper highlights as a major success of incorporating TabPFN.

Mira: They also note that this model reproduces the systematic evolution of alpha-decay half-lives along different isotopic chains and captures the shell-closure effect around N = one hundred twenty-six.

Lev: So it’s not just about getting a better average number; they’re showing that the model understands how things change systematically across different isotopes, like those sudden drops in decay rates you see near certain magic numbers.

Conclusion: Kai: So to wrap up this paper on the "Systematic Study of Proton, Two-Proton, Alpha, and Cluster Radioactivity Half-Lives based on the Deformed Gamow-like Model and Tabular Prior-data Fitted Network (TabPFN)", it seems the combination of DGLMb and TabPFN is a very solid way to handle multiple decay modes.

Mira: It provides a unified framework that keeps the physical structure of the original Deformed Gamow-like Model while using AI to get high-precision predictions on tabular data.

Lev: And they show good generalization in superheavy regions too, which is important when you're trying to predict things we haven't actually measured yet.

Kai: The final result is that the DGLMb plus TabPFN model with a sigma RMS of zero point four two three is the best of the four models they tested, and it doesn't show significant signs of overfitting when you look at their training and test set RMSEs <ref:2609.21646#pg1>.

Mira: It also keeps things physically interpretable because we can still see how the parameters like r0 and h affect the decay rates.

Lev: For me, the main thing is that this hybrid approach gives us a promising tool for describing multiple radioactive-decay modes, which is exactly what we need when trying to predict unknown nuclei.

College of Physics and Technology, Kunming University · School of Physics and Information Engineering, Zhaotong University · Key Laboratory of Economic System Simulation of Guizhou Province, Guizhou University of Finance and Economics

nucl-th

Submitted: 2026-09-18

Updated: 2026-10-29

Comments: 15 pages, 9 figure, 8 tables

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 83/100

The gist: A hybrid framework combining the deformed Gamow-like model (DGLM) with the Tabular Prior-data Fitted Network (TabPFN) is developed to improve half-life predictions for two-proton emission…

Key concepts

Deformed Gamow-like Model (DGLM)
This framework relates the half-life of a decay to its decay constant through a specific formula. It calculates the total penetration probability by averaging it over all possible directions, and defines the potential energy barrier between the emitted particle and the daughter nucleus.
Tabular Prior-data Fitted Network (TabPFN)
This is a machine learning component trained using structural parameters and decay parameters of radionuclides as input features. It learns conditional priors from synthetic tasks, allowing for efficient, task-agnostic predictions without needing complex optimization methods.
Penetration Probability (P)
This is the probability that a particle emitted from the nucleus successfully tunnels through the potential barrier to escape. The DGLM framework calculates this by averaging the penetration probability across all possible emission angles.
Shell-Closure Effect (N=126)
This refers to a specific stability feature in nuclei around neutron number 126. The model successfully reproduces the systematic evolution of alpha-decay half-lives and captures the pronounced variation associated with this shell closure, which is crucial for accurate predictions.

Terminology

Summary

A hybrid framework combining the deformed Gamow-like model (DGLM) with the Tabular Prior-data Fitted Network (TabPFN) is developed to improve half-life predictions for two-proton emission, proton emission, α decay, and cluster radioactivity

Model Performance Comparison

The DGLMb + TabPFN model achieves the best overall performance with a sigma RMS of 0.423, corresponding to an improvement of approximately 82. p. 0

Where the paper says it

Among the four considered models, DGLMb + TabPFN achieves the best overall performance, with σRMS = 0. 423, corresponding to an improvement of approximately 82.

2% over DGLMb. p. 0
Where the paper says it

A total of 583 radioactive nuclei are investigated, including 17 two-proton emitters, 42 proton emitters, 498 α emitters, and 26 cluster emitters. Among the four considered models, DGLMb + TabPFN achieves the best overall performance, with σRMS = 0.

After introducing TabPFN, the prediction errors for proton emission and α decay are reduced by approximately 80. p. 0
Where the paper says it

The model parameters are optimized for each decay mode using the least-squares method. After introducing TabPFN, the prediction errors for proton emission and α decay are reduced by approximately 80.

6% and 87.6%, respectively. p. 0
Where the paper says it

After introducing TabPFN, the prediction errors for proton emission and α decay are reduced by approximately 80. 6%, respectively.

p. 0
Where the paper says it

After introducing TabPFN, the prediction errors for proton emission and α decay are reduced by approximately 80. 6%, respectively.

The model also reproduces the systematic evolution of α-decay half-lives and the shell-closure effect around N = 126. p. 0
Where the paper says it

240, indicating good generalization capability without evident overfitting. The model also reproduces the systematic evolution of α-decay half-lives and the shell-closure effect around N = 126.

Theoretical Frameworks

The DGLM framework relates the half-lives T1/2 to the decay constant λ through T1/2 = ln 2 / λ × 10h. The total penetration probability P is obtained by averaging the penetration probability Pθ over all directions. p. 1

Where the paper says it

1 and 2 and the experimental half-life T exp 1/2 , the following expression is obtained T cal 1/2 T exp 1/2 = S expνP ScalνP = S exp Scal , (3) Within the unified decay theory framework, setting the theoretical preformation factor S cal to unity allows the experimental preformation factor to be extracted as[3] S exp = T cal 1/2 T exp 1/2 , (4) The assault frequency ν, which is related to the harmonic -oscillator frequency in the Nilsson potential [28], can be expressed as hν = ℏω ≈ 41 A1/3 MeV, (5) where h, ℏ, ω, and A denote the Planck constant, the reduced Planck constant, the angular frequency, and the mass number of the parent nucleus, respectively. In the deformed Gamow-like model, the total penetration probability P is obtained by averaging the penetration probability Pθ over all directions[29].

The total potential V (r, θ) between the emitted particle and the daughter nucleus can be expressed as V (r, θ) = (-V0, 0 ≤ r ≤ Rin(θ), VC (r, θ) + Vl(r), r > Rin(θ)>. p. 2
Where the paper says it

Yλ0(θ) denotes the spherical harmonic function. The total potential V (r, θ) between the emitted particle and the daughter nucleus can be expressed as V (r, θ) = ( −V0, 0 ≤ r ≤ Rin(θ), VC (r, θ) + Vl(r), r > Rin(θ).

Training and Optimization Strategy

The TabPFN is trained by using the structural parameters and decay parameters of the radionuclide to form the feature vector xi, and commonly used logarithmic values of preformation factors extracted from experiments are used as continuous regression labels. The model is developed by first learning conditional priors through a wide range of synthetic supervised tasks, achieving efficient and task-agnostic predictions without the need for optimization. p. 2

Where the paper says it

However, TabPFN has recently emerged as a powerful foundational model that offers outstanding predictive performance and real-time inference capabilities[32]. By learning conditional priors through a wide range of synthetic supervised tasks, TabPFN is able to generalize the learning process, achieving efficient and task-agnostic predictions without the need for optimization[32].

The optimal parameters r0 and h for each decay mode are determined using the least-squares method. p. 1
Where the paper says it

Part 3 conducts data analysis for different decay modes, focusing on the applicability of the DGLMb + TabPFN model. The optimal parameters r0 and h for each decay mode are determined using the least squares method, and the optimized models are used to predict the half-lives of relevant nuclides, which are then compared with existing empirical formulas.

Key Results Across Decay Modes

The optimal r0 values for two-proton emission, proton emission, α decay, and cluster emission are 1.24, 1.28, 1.30, and 1.27 fm respectively. p. 0

Where the paper says it

After introducing TabPFN, the prediction errors for proton emission and α decay are reduced by approximately 80. 6%, respectively.

After incorporating TabPFN, the predictive accuracy is substantially improved for the different decay modes. p. 1
Where the paper says it

This work combines the DGLM with TabPFN to improve the prediction accuracy of nuclide half-lives under different radioactive decay modes and to test the applicability of this method. The second section introduces the theoretical framework for combining the DGLM with TabPFN, presents prediction methods for nuclide decay half-lives using four combined models, and evaluates the predictive performance of each model using metrics such as standard deviation.

The σRMS values for proton emission and α decay are reduced by approximately 80. p. 0
Where the paper says it

The model parameters are optimized for each decay mode using the least-squares method. After introducing TabPFN, the prediction errors for proton emission and α decay are reduced by approximately 80.

6% and 87.6%, respectively. The DGLMb + TabPFN model also reproduces the systematic evolution of α-decay half-lives along different isotopic chains and successfully captures the pronounced variation associated with the N = 126 shell closure. p. 7
Where the paper says it

In particular, near the shell closure at N = 126, where half-lives exhibit distinct discontinuities, the DGLMb + TabPFN model successfully reproduces the variations observed in the experimental data. This indicates that the incorporation of T abP F N significantly improves the DGLMb model’s ability to predict α-decay half-lives across different isotope chains.

The DGLMb + TabPFN model performs best among the four models with σrms = 0. p. 3
Where the paper says it

direct learning of the pre-formed factors. simultaneously, DGLMb + TabPFN, constructed based on parameter optimization, performs best among the four models with σrms = 0.

423, indicating its superior predictive capability for radionuclide decay half-lives. p. 3
Where the paper says it

simultaneously, DGLMb + TabPFN, constructed based on parameter optimization, performs best among the four models with σrms = 0. 423, indicating its superior predictive capability for radionuclide decay half-lives.

p. 3
Where the paper says it

simultaneously, DGLMb + TabPFN, constructed based on parameter optimization, performs best among the four models with σrms = 0. 423, indicating its superior predictive capability for radionuclide decay half-lives.

p. 3
Where the paper says it

simultaneously, DGLMb + TabPFN, constructed based on parameter optimization, performs best among the four models with σrms = 0. 423, indicating its superior predictive capability for radionuclide decay half-lives.

Extrapolation and Generalization

The model was further applied to the heavier Z = 117, 118, 119, and 120 isotope chains and compared with results from the UDL formula to test the extrapolation capability of the DGLMb + TabPFN model in the superheavy nucleus region. The model possesses good extrapolation stability and predictive capability in the region of unknown superheavy nuclei. p. 8

Where the paper says it

Since no explicit shell correction terms were included in the model inputs, this suggests that DGLMb + TabPFN may have implicitly learned information related to shell structure from the training data. Although some discrepancies remain between the two methods near certain local extrema of highly neutron-rich nuclei, their overall trends and key inflection points show good consistency, indicating that the model possesses good extrapolation stability and predictive capability in the region of unknown superheavy nuclei.

The DGLMb + TabPFN model provides reasonable predictions for cluster radioactive half-lives and exhibits good stability across different emitter cluster systems, further validating the model’s feasibility for predicting unknown cluster radioactive systems. p. 9
Where the paper says it

The first and second columns represent the parent nucleus and the emitted cluster, respectively; the third column is the decay energy Qc; the fourth column is the orbital angular momentum quantum number ℓ; and the last three columns are, respectively, log10 T1/2(s) calculated using the DGLMb + TabPFN and UDL models, respectively. As can be seen from Table IX, the DGLMb + TabPFN model provides reasonable predictions for cluster radioactive half-lives and exhibits good stability across different emitter cluster systems, further validating the model’s feasibility for predicting unknown cluster radioactive systems.

The overall combination of DGLM and TabPFN significantly improves half-life predictions while retaining the physical interpretability of the conventional DGLM framework. p. 0
Where the paper says it

The model also reproduces the systematic evolution of α-decay half-lives and the shell-closure effect around N = 126. These results demonstrate that combining DGLM with TabPFN significantly improves the accuracy and robustness of radioactive-decay half-life predictions while retaining the physical interpretability of the original model.

p. 0
Where the paper says it

The model also reproduces the systematic evolution of α-decay half-lives and the shell-closure effect around N = 126. These results demonstrate that combining DGLM with TabPFN significantly improves the accuracy and robustness of radioactive-decay half-life predictions while retaining the physical interpretability of the original model.

p. 0
Where the paper says it

The model also reproduces the systematic evolution of α-decay half-lives and the shell-closure effect around N = 126. These results demonstrate that combining DGLM with TabPFN significantly improves the accuracy and robustness of radioactive-decay half-life predictions while retaining the physical interpretability of the original model.

Parameter Fitting Details

For α decay, the optimal hindrance factors are hEO = hOE = 0.220 and hOO = 0. p. 7

Where the paper says it

220, hOE = 0. 220, hOO = 0.

510, while h = 1. p. 3
Where the paper says it

0 Test data (RMSD) 0. 510 Train mean = 0.

18 is obtained for cluster emission. p. 6
Where the paper says it

Experimental data for the proton-emission half-lives, spin, and parity are taken from Ref. The Qp values are obtained from Eq.

The training set RMSE for all 498 kernels was 0. p. 7
Where the paper says it

This table is intended to evaluate both the model’s fitting accuracy and its generalization ability. For all 498 kernels, the model’s training, test, and overall RMSEs were 0.

208, the test set RMSE was 0. p. 7
Where the paper says it

195, suggesting that the model describes the even-even kernel most accurately. The test set RMSE for odd–even nuclei was 0.

305, and the overall RMSE was 0. p. 0
Where the paper says it

For α decay, the training, test, and overall RMSEs are 0. 305, and 0. p. 0

Where the paper says it

For α decay, the training, test, and overall RMSEs are 0. 305, and 0.

240. The relatively limited difference between the training and test errors indicates that the DGLMb + TabPFN model has good generalization capability for α-decay predictions and shows no evident signs of overfitting. p. 7
Where the paper says it

376, respectively—relatively high values indicating that the presence of unpaired nucleons increases the complexity of α decay patterns. Overall, the difference in error between the training and test sets is limited, indicating that the DGLMb+TabPFN model exhibits good stability and does not show significant overfitting.

The cluster decay half-lives are well described by the DGLMb + TabPFN model across the lead region. p. 9
Where the paper says it

The last three columns present the theoretical halflives calculated using the DGLMb + TabPFN model, the DGLMb model, and the UDL empirical formula, respectively. As shown in Table VIII, the DGLMb + TabPFN model provides a good description of the cluster radiation halflife across the lead region and offers higher predictive reliability compared to the DGLMb model and the UDL empirical formula.

The training, test, and overall RMSEs for the 498 α-decay nuclei are 0. p. 7
Where the paper says it

This table is intended to evaluate both the model’s fitting accuracy and its generalization ability. For all 498 kernels, the model’s training, test, and overall RMSEs were 0.

208, 0.305, and 0.240 respectively. p. 7
Where the paper says it

276 All nuclei 498 0. 240 Subsequently, the hindrance factor h was fitted using experimental data on ground-state α decay from 142 even-Z odd-N nuclei, 119 odd-Z even-N nuclei, and 65 doubly odd nuclei, respectively.

Empirical Formula Comparison

The theoretical half-lives of p and 2p radioactivity were calculated using the corresponding GNL formulas, while the theoretical half-lives of α decay and cluster radioactivity were uniformly calculated using the UDL. The predictions from DGLMb + TabPFN are generally close to those from GNL, indicating good agreement between the two methods regarding the half-lives of these candidate nuclei. p. 6

Where the paper says it

In this table, the first three columns represent, respectively, the nuclides of the possible proton-emitting radionuclides, the decay energy Qp, and the orbital angular momentum l. It can be seen that the predictions from DGLMb+TabPFN are generally close to those from GNL, indicating good agreement between the two methods regarding the half-lives of these candidate nuclei.

The DGLMb + TabPFN model provides a good description of the cluster radiation halflife across the lead region and offers higher predictive reliability compared to the DGLMb model and the UDL empirical formula. p. 9
Where the paper says it

The last three columns present the theoretical halflives calculated using the DGLMb + TabPFN model, the DGLMb model, and the UDL empirical formula, respectively. As shown in Table VIII, the DGLMb + TabPFN model provides a good description of the cluster radiation halflife across the lead region and offers higher predictive reliability compared to the DGLMb model and the UDL empirical formula.

The DGLMb + TabPFN model successfully reproduces the variations observed in experimental data near N = 126, indicating that it significantly improves the ability to predict α-decay half-lives across different isotope chains. p. 7
Where the paper says it

In particular, near the shell closure at N = 126, where half-lives exhibit distinct discontinuities, the DGLMb + TabPFN model successfully reproduces the variations observed in the experimental data. This indicates that the incorporation of T abP F N significantly improves the DGLMb model’s ability to predict α-decay half-lives across different isotope chains.

The DGLMb + TabPFN model provides a promising tool for describing multiple radioactive-decay modes and for predicting the decay properties of experimentally unknown nuclei. p. 9
Where the paper says it

The first and second columns represent the parent nucleus and the emitted cluster, respectively; the third column is the decay energy Qc; the fourth column is the orbital angular momentum quantum number ℓ; and the last three columns are, respectively, log10 T1/2(s) calculated using the DGLMb + TabPFN and UDL models, respectively. As can be seen from Table IX, the DGLMb + TabPFN model provides reasonable predictions for cluster radioactive half-lives and exhibits good stability across different emitter cluster systems, further validating the model’s feasibility for predicting unknown cluster radioactive systems.

The present hybrid approach therefore provides a promising tool for describing multiple radioactive-decay modes and for predicting the decay properties of experimentally unknown nuclei. p. 9
Where the paper says it

The first and second columns represent the parent nucleus and the emitted cluster, respectively; the third column is the decay energy Qc; the fourth column is the orbital angular momentum quantum number ℓ; and the last three columns are, respectively, log10 T1/2(s) calculated using the DGLMb + TabPFN and UDL models, respectively. As can be seen from Table IX, the DGLMb + TabPFN model provides reasonable predictions for cluster radioactive half-lives and exhibits good stability across different emitter cluster systems, further validating the model’s feasibility for predicting unknown cluster radioactive systems.

p. 9
Where the paper says it

The first and second columns represent the parent nucleus and the emitted cluster, respectively; the third column is the decay energy Qc; the fourth column is the orbital angular momentum quantum number ℓ; and the last three columns are, respectively, log10 T1/2(s) calculated using the DGLMb + TabPFN and UDL models, respectively. As can be seen from Table IX, the DGLMb + TabPFN model provides reasonable predictions for cluster radioactive half-lives and exhibits good stability across different emitter cluster systems, further validating the model’s feasibility for predicting unknown cluster radioactive systems.

p. 9
Where the paper says it

The first and second columns represent the parent nucleus and the emitted cluster, respectively; the third column is the decay energy Qc; the fourth column is the orbital angular momentum quantum number ℓ; and the last three columns are, respectively, log10 T1/2(s) calculated using the DGLMb + TabPFN and UDL models, respectively. As can be seen from Table IX, the DGLMb + TabPFN model provides reasonable predictions for cluster radioactive half-lives and exhibits good stability across different emitter cluster systems, further validating the model’s feasibility for predicting unknown cluster radioactive systems.

p. 9
Where the paper says it

The first and second columns represent the parent nucleus and the emitted cluster, respectively; the third column is the decay energy Qc; the fourth column is the orbital angular momentum quantum number ℓ; and the last three columns are, respectively, log10 T1/2(s) calculated using the DGLMb + TabPFN and UDL models, respectively. As can be seen from Table IX, the DGLMb + TabPFN model provides reasonable predictions for cluster radioactive half-lives and exhibits good stability across different emitter cluster systems, further validating the model’s feasibility for predicting unknown cluster radioactive systems.

Summary of Findings

The work concludes that the combination of DGLM and TabPFN significantly improves half-life predictions while retaining the physical interpretability of the conventional DGLM framework. The present hybrid approach therefore provides a promising tool for describing multiple radioactive-decay modes and for predicting the decay properties of experimentally unknown nuclei. The final results indicate that the accuracy of the model was significantly improved by introducing TabPFN to enable in0. p. 3

Where the paper says it

423, respectively. The results indicate that the accuracy of the model was significantly improved by introducing TabPFN to enable in0.

p. 3
Where the paper says it

423, respectively. The results indicate that the accuracy of the model was significantly improved by introducing TabPFN to enable in0.

p. 3
Where the paper says it

423, respectively. The results indicate that the accuracy of the model was significantly improved by introducing TabPFN to enable in0.

25 0.30 0.35 0.40 0.45 0.50 Training data (RMSD). p. 3
Where the paper says it

The results indicate that the accuracy of the model was significantly improved by introducing TabPFN to enable in0. 50 Training data (RMSD) 0.

p. 3
Where the paper says it

The results indicate that the accuracy of the model was significantly improved by introducing TabPFN to enable in0. 50 Training data (RMSD) 0.

The DGLMb + TabPFN model, constructed based on parameter optimization, performs best among the four models with σrms = 0. p. 3
Where the paper says it

direct learning of the pre-formed factors. simultaneously, DGLMb + TabPFN, constructed based on parameter optimization, performs best among the four models with σrms = 0.

p. 3
Where the paper says it

direct learning of the pre-formed factors. simultaneously, DGLMb + TabPFN, constructed based on parameter optimization, performs best among the four models with σrms = 0.

423, indicating its superior predictive capability for radionuclide decay half-lives. The model exhibits good stability and does not show significant overfitting. p. 7
Where the paper says it

376, respectively—relatively high values indicating that the presence of unpaired nucleons increases the complexity of α decay patterns. Overall, the difference in error between the training and test sets is limited, indicating that the DGLMb+TabPFN model exhibits good stability and does not show significant overfitting.

p. 7
Where the paper says it

376, respectively—relatively high values indicating that the presence of unpaired nucleons increases the complexity of α decay patterns. Overall, the difference in error between the training and test sets is limited, indicating that the DGLMb+TabPFN model exhibits good stability and does not show significant overfitting.

The DGLMb + TabPFN model provides predictions that are much closer to the experimental α-decay half-lives. p. 7
Where the paper says it

247 after incorporating TabPFN, corresponding to a reduction of approximately87. This significant decrease indicates that theDGLMb + TabPFN model provides predictions that are much closer to the experimental α-decay half-lives.

p. 7
Where the paper says it

247 after incorporating TabPFN, corresponding to a reduction of approximately87. This significant decrease indicates that theDGLMb + TabPFN model provides predictions that are much closer to the experimental α-decay half-lives.

The incorporation of TabPFN substantially improves the predictive accuracy of the DGLM model for α decay. p. 7
Where the paper says it

This significant decrease indicates that theDGLMb + TabPFN model provides predictions that are much closer to the experimental α-decay half-lives. Therefore, the incorporation of T abP F Nsubstantially improves the predictive accuracy of the DGLM model for α decay.

p. 7
Where the paper says it

This significant decrease indicates that theDGLMb + TabPFN model provides predictions that are much closer to the experimental α-decay half-lives. Therefore, the incorporation of T abP F Nsubstantially improves the predictive accuracy of the DGLM model for α decay.

The DGLMb + TabPFN model successfully reproduces the variations observed in experimental data near N = 126, indicating that it significantly improves the DGLMb model’s ability to predict α-decay half-lives across different isotope chains. p. 7
Where the paper says it

In particular, near the shell closure at N = 126, where half-lives exhibit distinct discontinuities, the DGLMb + TabPFN model successfully reproduces the variations observed in the experimental data. This indicates that the incorporation of T abP F N significantly improves the DGLMb model’s ability to predict α-decay half-lives across different isotope chains.

p. 7
Where the paper says it

In particular, near the shell closure at N = 126, where half-lives exhibit distinct discontinuities, the DGLMb + TabPFN model successfully reproduces the variations observed in the experimental data. This indicates that the incorporation of T abP F N significantly improves the DGLMb model’s ability to predict α-decay half-lives across different isotope chains.

The overall combination of DGLM and TabPFN significantly improves half-life predictions while retaining the physical interpretability of the conventional DGLM framework. The present hybrid approach therefore provides a promising tool for describing multiple radioactive-decay modes and for predicting the decay properties of experimentally unknown nuclei. The final results indicate that the accuracy of the model was significantly improved by introducing TabPFN to enable in0.25 0.30 0.35 0.40 0.45 0.50 Training data (RMSD). The DGLMb + TabPFN model, constructed based on parameter optimization, performs best among the four models with σrms = 0.423, indicating its superior predictive capability for radionuclide decay half-lives. The model exhibits good stability and does not show significant overfitting. The DGLMb + TabPFN model provides predictions that are much closer to the experimental α-decay half-lives. The incorporation of TabPFN substantially improves the predictive accuracy of the DGLM model for α decay. The DGLMb + TabPFN model successfully reproduces the variations observed in experimental data near N = 126, indicating that it significantly improves the DGLMb model’s ability to predict α-decay half-lives across different isotope chains. The overall combination of DGLM and TabPFN significantly improves half-life predictions while retaining the physical interpretability of the conventional DGLM framework. The present hybrid approach therefore provides a promising tool for describing multiple radioactive-decay modes and for predicting the decay properties of experimentally unknown nuclei. The final results indicate that the accuracy of the model was significantly improved by introducing TabPFN to enable in0.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improve half-life prediction accuracy for proton emission by 80.6% by incorporating TabPFN into the DGLMb model, allowing it to correct the systematic bias of DGLMb model’s predictions and bring them into closer alignment with experimental data.

  2. Enhance α decay prediction robustness by reducing the root-mean-square deviation from 1.999 to 0.247 after incorporating TabPFN, which significantly improves accuracy and shows no evident signs of overfitting.

  3. Enable extrapolation for unknown nuclides by utilizing the DGLMb + TabPFN model, as it provides good extrapolation stability and predictive capability in the region of unknown superheavy nuclei when compared with UDL formula results.

  4. Improve cluster radioactivity predictions by optimizing nuclear radius parameter r0 to 1.27 fm and hindrance factor h to 1.18 for odd-A nuclei, leading to a model that provides a good description of the cluster radiation halflife across the lead region.

  5. Develop a unified framework capable of predicting four different charged-particle emission decay processes—two-proton emission, proton emission, alpha decay, and cluster radioactivity—using the combination of DGLM and TabPFN.

Abstract

A hybrid framework combining the deformed Gamow-like model (DGLM) with the Tabular Prior-data Fitted Network (TabPFN) is developed to improve half-life predictions for two-proton emission, proton emission, α decay, and cluster radioactivity. A total of 583 radioactive nuclei are investigated, including 17 two-proton emitters, 42 proton emitters, 498 α emitters, and 26 cluster emitters. Among the four considered models, DGLM b+ TabPFN achieves the best overall performance, with σ RMS=0.423, corresponding to an improvement of approximately 82.2% over DGLM b. The model parameters are optimized for each decay mode using the least-squares method. After introducing TabPFN, the prediction errors for proton emission and α decay are reduced by approximately 80.6% and 87.6%, respectively. For α decay, the training, test, and overall RMSEs are 0.208, 0.305, and 0.240, indicating good generalization capability without evident overfitting. The model also reproduces the systematic evolution of α-decay half-lives and the shell-closure effect around N=126. These results demonstrate that combining DGLM with TabPFN significantly improves the accuracy and robustness of radioactive-decay half-life predictions while retaining the physical interpretability of the original model.

Sources

Related papers